Description of problem: While doing U3 errata testing for the newly added gnome-python2-applet, executing the testcode causes a seg fault both remote X and locally. Version-Release number of selected component (if applicable): gnome-python2-applet-1.99.14-5 How reproducible: Always Steps to Reproduce: 1. log into the desktop and open a terminal. 2. execute the following testcode: % python >>> import gtk >>> import gnome.applet >>> gnome.applet.Applet() 3. Actual results: Seg Fault Expected results: nothing blows up. applet created. Additional info:
Created attachment 103035 [details] backtrace from gnome-python2-applet Backtrace from i386
Just for sanity sake -- you're running on a box with a valid DISPLAY that's running a GNOME desktop, yes? (And more preferably, a RHEL3 gnome desktop)
Yes, this fails on the stable systems in RDU remotely and I had dkl test them locally. It seg faults in either case. Other apps forward and come up correctly for me here (mozilla, xterm, gedit).
Yeah, but applets are "special" since they depend on more than just a display. So if you're not running it all on the same machine (ie, the whole desktop run from one box), I wouldn't be surprised if it failed. (Discounting doing wacky things like turning on ORBit over tcp/ip)
Looks like the problem was with the test case. older gnome-python weren't robust against instantiation w/o a bonobo factory to go with them. It seems to work fine w/ a normal applet, but not the given test case